3038 Inovgnaic Chemistvy, Vol. 13, No. 12, 19 74 characteristics of the substituents have been attempted: Hammett crP constants: substituent electronegativities from vm bond vibrations: empirically derived mutually consistent group electronegativities using various data: Hammett uI constants,6 and substituent steric requirements. The advantages and disadvantages of these materials in looking for such correlation phenomena have been recently considered in some We find only a weak positive correlation of the IS with the substituent electronegativities indicating a slightly increasing total s-electron density at the tin nucleus with increasing electronwithdrawiiig capacity of the para substituent. In addition, we find some correlation of the IS with the QS as the magnitude of the observed splitting generally decreases with increasing shift values? Such behavior is consistent with slightly decreasing coordination abilities of the nitrogen atoms toward the central tin atom? Both of the above correlations would be consistent with the conclusion that the degree of coordination between the tin and the nitrogens in these compounds varies due to the inductive transmitta!. of electronic effects from the phenyl para positions. Qne might thus expect to find a slightly larger central hole in the F-substituted compound than in the parent ClzSnTPP. Unfortunately, only the structure of the latter has been reported.” All seven compounds studied here exhibit distinct asymmetries in their doublet Mossbauer spectra. The ratio of the high-energy wing area to the lowenergy wing is indicated for each sample spectrum in Table I . It is quite apparent that increased area asymmetries are accompanied by decreased splittings. Such asyinmetry variations (if indeed due t o variations in the bonding environment of the tin) should represent variations in the spatial anisotropy of the recoil-free fraction of the tin atoms.” One might expect the tin to be more tightly bound in the “plane” roughly defined b y the nitrogen atoms rather than along the C1-SnCl axis. All of the Mossbauer measurements were obtained with finely powdered samples but, because nonrandom orientations of the microcrystals could also lead t o spectral asymmetries, several of the compounds were repeatedly reground and reexamined. No significant changes in the doublet asymmetries were observed, thus indicating a random orientation of the polycrystalline absorbers. Possible line shape model errors in the spectrum-fitting routine have been considered and we are quite confident that they are negligible in the determination of the IS and QS from these well-fitted spectra. The distribution of the calculated residuals is uniform across the absorption peak of each spectrum. The supposed quadrupolar-split spectra may, however, actually be the result of two very closely spaced singlets originating from two crystal forms of each compound; the two forms differ, perhaps, by the orientations of the phenyl groups, We have no evidence t o support such a supposition and these data indicate that both crystal forms would he equaliy prevalent which on stereochemical considerations o f the possible structures would be unlikely. Careful analyses of these spectra in a strong external magnetic (4) H. H. Jaffe, C h e m Rev., 53, 191 (1953). (5) P. R. Wells, Progu. Phys. Ovg. Chem., 6, 1 1 1 (1968). (6) R. W.Taft, J. Chem. Phys., 64, 1809 (1960); P. R. Wells, Chem. Rev., 63, 171 (1963). (7) A. D. Adler, A n n . N . Y.k c a d . Sci., 206, l(l973). (8) M. Meat-Ner and A. D. Adler, J. Amer. Chem. SOC.,94,4763 (1972). (9) R. L. Collins and J. C. Travis, i$/lossbauerEff. Methodol., 3, 123 (1967). (10)J. L. Hoard, Science, 194, 1295 (1971);D. M. Collins, W, R . Scheidt, and J. L. Hoard, J. Amev. Chem. SOC.,94,6689 (1992). (1 I ) S. V. Karyagin, Dokl. Akad. Nauk S S S R , 3.40, 1 1 0 (1963).
Notes field would, however, differentiate between a genuine doublet and two closely spaced singlets. Chemical impurities at a 50% level are known not to be significant in these chromatographically purified and analytically characterized compounds.’ We conclude, therefore, that though all of these complexes exhibit discernible doublet 119mSnMossbauer spectra with various magnitudes of the splitting, the causes are not completely clear. The doublets might be two singlets, closely spaced and representing two structural forms of the complexes, although we consider this unlikely. Differences in the crystal packing arrangements of these solid-state complexes might also yield variations in the electric field gradient (EFG) at the tin nuclear site, but such sensitivity of the EFG to long-range interactions would be highly unusual for such molecular solids. As a further consideration, we have estimated the effect of slight movements of the tin atom along the Cl-Sn-421 axis on the basis of the point-charge model” and found that if the atom were t o pop out of the central hole by up to 0.2 A, then the QS might be expected to vary b y about lO-15%, as observed. The effect of such a distortion upon the IS would probably be very small as the accompanying rehybridization of the tin bonding orbitals would not substantially affect the total s-electron density at the tin nucleus. However, an independent test (and confirmation) of this explanation of the data requires further X-ray structural analyses of some of the substituted complexes. ~
c
~ Portions ~ of this~work were ~ carried~ out
esearch Council-National Bureau of Standards Postdoctoral Research Associate and this support is gratefully acknowledged. Regis%PyNoe C1, SnT(p-i-C,H,)PP, 52628-87-2; 61,SnT(pCB,)PP, 26334-83-8; Cl,§nTPP, 26334-85-0; C1, SnT(p-OCA,)PP, 26334-82-7; CI,SnT(p-Br)PP, 52628-86-1; C1, SnT(p-CI)PP, 2 4 3 3 4 84-9; Cl,SnT(p-F)PP, 52628-88-3. (32) B. W. Fitzsimmons, N.J. Seeley, and A. W . Smith, Chem. Cornmun., 390 (1968); R. V. Parish and R. PI. Platt, Inoug. Chim. Acta, 4, 6 5 (1970);N. W. G. Debye and J . J. Zuckerman in “Deter. mination of Organic Structures b y Physical Methods,” Vol. 5, P.C. Nachod and J. J. Zuckerman, Ed., Academic Press, New York and London, 1973,p 235.
Contribution from the Department of Chemistry, University of Georgia, Athens, Georgia 30602
tudiss on Tungsten CP R. B. King,* M. S. Saran,’ and S. P. Anand, Received June d 4, 19 74
ATC403 8 3 0
Cleavage of the metal-metal bond in bimetallic metal carbonyl derivatives such as [CsHsM(@O)& (M = Mo, W), M2e), and [C5B,M(CO),], (M = Fe, Ru) by reagents such as halogens and alltali metals provides useful routes to interesting monometallic transition metal derivatives.3 In this connection, the recently reported4 unsymmetrical bimetallic tungsten car bony1 nitrosyl hydride (1) Post doctoral research associate, 1969-1 9 74. (2) Postdoctoral research associate, 19’/3-997dL (3) K. B~ King, ‘“‘TransitionMetal Organometallic Chemistry: An Introduction,’’ Academic Press, New York, N. Y., 1969. (4)M.Andrews, D. L. Tipton, 9. W. Kirtley, and Ea. Bau, J Chem. Soc., Chern, Commtcn., 181 (6975).
~
~
Inorganic Chemistvy, VoZ. 13,No. 12, 1974 3039
Notes HW2(CO)9N0 (I) was of interest since similar tungsten-tungsten bond cleavage reactions could give W(CO),(NO)X derivatives isoelectronic with the well-known3 M(C0)5X (M = Mn, Tc, Re) derivatives. These considerations led to an investigation of the reactions of HW2(C0)9N0 with sodium amalgam and with halogenating reagents such as nitrosyl chloride and iodine. In our hands, reactions of HW2(CO)9N0 with sodium amalgam in tetrahydrofuran solution followed hy addition of methyl iodide or trimethyltin chloride gave only traces of yellow rather intractable products without the v(N0) frequency in their infrared spectra expected for a RW(CO),NO derivative. However, halogenations of HW2(CO)9N0 with nitrosyl chloride and with iodine provided novel synthetic routes to the halides trans-W(CO),(NO)X (11, X = C1, I). After our work o n the syntheses of the trans-W(CO),(NO)X halides from HW2(CO)9N0 was essentially completed, a report b y Barraclough, Bowden, Colton, and Commons5 on the syntheses of the same trans-W(@O),(NO)X halides by nitrosation of [(C2H5>,N][W(CO),X] became available to us. We have therefore phased out our work in this area. This note reports our syntheses of trans-W(CO),(NO)X (11, X = C1, I) from HW2(CO)9N0 as well as some reactions of trans-W(CO),(N0)I which are not discussed in recent papers b y Colton and common^.^.^
I
I1
Experimental Section Commercial hexacarbonyltungsten (Pressure Chemical Corp., Pittsburgh, Pa.), after resublimation, was converted to HW, (CO),NO by reaction with NaBH, in tetrahydrofuran to give Na[HW,(CO),,] followed by nitrosation of this sodium salt with sodium nitrite and phosphoric acid in aqueous s ~ l u t i o n . ~ "From 50 g (142 mmol) of W(CO), this procedure gave 17 g (37% yield) of HW,(CO),NO after crystallization from dichloromethane and heating at -45" (1 mm) to remove any unreacted W(CO), . tert-Butyl isocyanide was prepared from fert-butylamine via tert-b~tylformamide.~Other aspects of the experimental procedures are similar to those described in previous papers from this laboratory.1u Reaction of HW,(CO),NO with Nitrosyl Chloride. Nitrosyl chloride was passed for 10 min through a solution of 0.73 g (1.12 mmol) of HW,(CO),NO in 75 ml of benzene. The color changed from red-orange to brown and a yellow precipitate separated. Removal of solvent at -25" (35 mm) followed by sublimation at 45-50" (1.5 mm) gave a 35% yield of a yellow sublimate of trans-W(CO),(NO)Cl identified by molecular weight and elemental analyses (C, N, C1, 0, W) and by comparison of its infrared spectrum in the v(C0) region with that reported in the literature.' Reaction of HW,(CO),NO with Iodine. A solution of 11.0 g (16.9 mmol) of HW,(CO),NO in 400 ml of dichloromethane was treated dropwise with a solution of 7.0 g (27.6 mmol as I,) of iodine in 250 ml of dichloromethane. 'The filtered dichloromethane solution was concentrated nearly to dryness and then chromatographed on a 3 X 50 cm Florisil column. A small amount of W(CO), was first eluted with pure hexane. The major yellow band of trans-W(CO),(NO)I was eluted with 1 :9 dichloromethane-hexane. Evaporation of this eluate gave 5.4 g (70% yield based on available nitrosyl groups) ( 5 ) C. G. Barraclough, J. A. Bowden, R. Colton, and C . J . Commons, Aust. J. Chem., 26, 241 (1973). (6) R. Colton and C. J . Commons, Aust. J. Chem., 26, 1487
(1 973). (7)R. Colton and C. J. Commons, Aust. J. Chem., 26, 1493 (1973). (8)J. P. Olsen, T. F. Koetzle, S. W. Kirtley, M. Andrews, D. L. Tipton, and R. Bau, J . Amer. Chem. Soc., 96,6621 (1974). (9) I. Ugi and R. Meyr, Chem. Ber., 93,239 (1960). ( I O ) R. B. King and M. N. Ackermann, Znorg. Chem., 13, 637 (1974).
of trans-W(CO),(NO)I, identified by molecular weight and elementidl analyses (C, N, I, 0, W) and comparison of its infrared spectrum in the v(C0) region with that reported in the literature.' Reactions of trans-W(CO),(NO)I with fertButy1 Isocyanide. ( a ) Hexane at Room Temperature. A mixture of 0.453 g (1 mmol) of trans-W(CO),(NO)I, 1 ml(0.72 g, 8.7 mmol) of terl-butyl isocyanide, and 60 ml of hexane was stirred for 18 hr at room temperature. A brown solid separated during the course of this reaction. Dichloromethane (25 ml) was added to dissolve this brown solid. Filtration and concentration of the filtrate to -25 ml at -25" (35 mm) gave 0.32 g (57% yield) of cis-(Me,CNC),W(CO),(NO)I. (b) Boiling Benzene (80"). A mixture of 0.453 g (1 mmol) of trans-W(CO),(NO)I, 1 ml (0.72 g, 8.7 mmol) of ierl-butyl isocyanide, and 60 ml of benzene was boiled under reflux for 2.5 hr. Removal of benzene at 25" (35 mm) was followed by chromatography on a 2.5 X 35 cm Florisil column to give 0.27 g (48% yield) of cis-(Me,CNC),W(CO),(NO)I, eluted with 2:3 dichloromethane-hexane, followed by 0.14 g (23% yield) of (Me,CNC),W(CO)(NO)l, eluted with pure dichloromethane. (c) Boiling Toluene (110"). A mixture of 0.453 g (1 mmol) of tmns-W(CO),(NO)I, 1.5 ml(l.08 g, 13.0 mmol) of terrt-butyl isocyanide, and 60 ml of toluene was boiled under reflux for 17 hr. Chromatography on a 2.5 X 35 cm Florisil column gave 0.28 g (42% yield) of bright yellow (Me,CNC),W(NO)I. Properties of the (Me,CNC),W(CO),-,(NO)I Derivatives. (a) cis-(Me,CNC),W(CO),(NO)I: yellow to orange, mp 139-140". Infrared spectrum (CH,Cl,): v(CN) 2201 (m) and 2182 (m) c m - ' ; v(C0) 2040 (s) and 1976 (s) cm-'; v(N0) 1642 (s) cm-'. Proton nmr (CDC1,): T 8.43. Anal. Calcd for C,,H,,IN,O,W: C, 25.6; H, 3.2; N, 7.4; 0, 8.5. Found: C, 25.3; H, 3.3; N, 7.4; 0, 9.2. (b) (Me,CNC),W(CO)(NO)I: yellow, dec pt 166". Infrared spectrum (CH,CI,): v(CN) 2189 (w) and 2139 (s) cm"; v(C0) 1961 (s) cm'l; u ( N 0 ) 1602 (s) cm-'. Proton nmr (CDC1,): P 8.44. Anal. Calcdfor C,,H,,IN,O,W: C, 31.1; H, 4.4; N, 9.1; 0 , 5 . 2 . Found: C, 30.9; H, 4.2; N, 8.9; 0, 5.8. (c) trans-(Me,CNC),W(NO)I: bright yellow, dec pt >185'. Infrared spectrum (CH,Cl,): v(CN) 2112 (s) and 2070 (w, sh) cm-'; no v(C0); v(N0) 1564 (m) cm". Proton nmr (CDCl,): r 8.49. Anal. Calcd for C,,H,,IN,OW: C, 35.7; H, 5.4; N, 10.4; 0 , 2.4. Found: C, 34.8; H, 5.3; N, 10.1; 0, 2.8. Preparation of [(C,H,),PCH,CH,P(C,H,),]W(CO),(NO)I. Reaction of 0.453 g (1 mmol) of trans-W(CO),(NO)I and 0.9 g (2 mmol) of (C,H,),PCH,CH,P(C,H,), in 70 ml of boiling hexane for 1 hr followed by chromatography on Florisil in hexane solution gave after elution with dichloromethane 0.72 g (91% yield) of yellow [(C,H,),PCH,CH,P(C,HS),]W(CO),(NO)I,mp 182-183" dec. lnfrarea spectrum (CH,Cl,): n(C0) 2030 (s) and 1950 (s) cm-'; v(N0) 1630 (m) cm-'. Anal. Calcd for C,,H,,INO,P,W: C, 42.3; H, 3.0; I , 16.0; N, 1.8; 0, 6.0. Found: C, 42.7; H, 3.3; I, 17.3; N, 1.7; Q6.1. Conversion of trans-W(CO),(NO)I to C,H,W(CO),NO. h mixture of 0.906 g (2 mmol) of trans-W(CO),(NO)I, 0.538 g (%mmol) of thallium cyclopentadienide, and 75 ml of tetrahydrofuran was boiled under reflux for 75 min. Tetrahydrofuran was then removed at -25' (35 mm). A solution of the residue in 100 ml of hexane was chromatographed on a 2.5 X 35 cm Florisil column. Elution of the yellow band with hexane followed by concentration of the hexane eluate to -10 ml and cooling to 0" gave 0.41 g (61% yield) of yellow-orange crystalline C,H,W(CO),NO, mp 106-107" (lit." mp 105-107"), identified by its infrared v(C0) and u(N0) frequencies. Use of sodium cyclopentadienide instead of thallium cyclopentadienide in the above reaction also gave C,H,W(CO),NO, but only in -22% yield. Similar reactions of trans-W(CO),(NO)I with phenyllithium in benzene, methyllithium in diethyl ether, allylmagnesium chloride in tetrahydrofuran, and K,C,H, in tetrahydrofuran on a 1mmol scale all gave negative results or intractable mixtures apparently not containing the desired organotungsten compounds.
Discussion The formation of trans-W(CO),(NO)Cl from the reaction o f HW2(CO)9N0 with nitrosyl chloride is rather unexpected since a similar reaction of W(CO)6 with nitrosyl chloride gives the carbonyl-free complex [W(NO)2Clz], . 1 2 This observation suggests that nitrosyl chloride only functions as a &lorina.ting agent in its reaction with HW,(CO),NO, which already (11) E. 0. Fischer, 0. Beckert, W. Hafner, and H.0. Stah'i, Z.Naturforsch. B , 10, 598 (1955). (12) F. A. Cotton and B. F. G. Johnson, Inorg. Chem., 3 , 1609
(1 9 64).
3840 Inorganic Chemistry, Vol. I d , No. 12, I974 has the necessary nitrosyl group to form trans-W(CO),(NO>Cl, whereas nitrosyl chloride must also function as a nitrosylating agent in its reaction with W(CO)6. This dichotomy in the action of nitrosyl chloride on transition metal systems can account for the very different pathways in its reactions with W(CO), and HW2(CO)9NO. The tendency for HW,(CO),NO to f5rm halides of lhe type traiz~-W(CO)~(EalO)X with halogenating agents is also demonstrated by the reaction of HW2(CO)gNO with. elemental iodine to give trans-W(CO),(IW)I in yields up to 70% based on available nitrosyl groups. Since only one of the two ~ has a nitrosyl ~ group, halo-~ tungsten atoms in ~ genation of W 2 ( C 0 ) 9 N 0 must also lead to other tungsten complexes from the tungsten atom without the nitrosyl group. These tungsten by-products can compiicate the separation of pure t r a ~ ~ ~ W ~ derivatives ~ ~ ~ , from ( ~ ~the) halo% genation of HW,(@0)9N0. In the preparation of Ivuns-W(CO),(NO)I from HW,(C0)9N0 and iodine, separation o f a pure product is facilitated if excess iodine is used. The rernaining tungsten products ? h mthis preparation can appear as brown solids insoluble in dichlorornethane or as a brown., very strongly adsorbed band on the chromatography column after removal of the trarans-W(CO)4(N.rO)I. The iodination o f HW2(CO)9N0 in benzene, furthermore, can give erratic yields of a blue volatile solid of stoichiometry [W(CO),l],; the nature of this product is uncertain. Reaction of MW,(CO)9N0 with bromine has been shown spectroscopically to produce trans-W(CQ)4(NQ)Br, but a reliable procedure has not. been found for separating pure trans-W(CO),(NO)Br from the W(C0)6 concurrently produced in this reaction. Colton and Commons have reported some reactions of tr~ns-w(Co),(NO)X(X zz C1, r, I) with simple tertiary phosphines and arsines6 to give either monosubstituted mer-W(CO)3L(NO)% or disubstituted C ~ ~ - W ( C O ) ~ L ~ ( N (E O )= % (C&),P, (C6H5)3As;X = C1, Br, 1) depending upon the severity of the reaction conditions. We have found conditions for the reaction of trans-W(CO),(NO)l with tert-butyl isocyanide where the dicarbonyl cis-(Me 3CNC)2W(C0)2(N6(B)I, the monocarbonyl (Me3CNC)3W(C0>(No)l,or even the carbonyl-free fruns-(~e3CNC>,W(No)Ican be obtained simply by varying the reaction temperature from room temperature to 110". Stereochemistrics of cis-(Me3CNe),W(eca>,(~Y))I with the pairs of both the carbonyl and fert-butyl isocyanide ligands in relative cis positions (ie., IHIa or IlIbj and of trans(Me3CNC)4W(NO)I with the nitrosyl and iodide ligands in trans positions ( i e . ?IV with local &h, symmetry) are supported by the observed numbers and positions of v(CN.>and v(CO) frequencies, but the available infrared data do not allow either a decision between IIIa and IBIb for the disubstituted derivative or an unequivocal elucidation of the slereochemistry of the trisubstituted derivative (Me3CNC)3W(CO)(N0)I. However, structure lIBa s e e m more probable than BIlb n
Notes for the disubstituted cis-(Me3@NC),~(GB),(NO))[in view of its single, sharp fert-butyl proton nmr resonance and its formation from traizs-W(CO)4(NO)I. In C ~ S - ( M ~ ~ C N C ) ~ W ( C O ) ~ (N0)P the two v ( C 0 ) frequencies are separated by 64. cm-' whereas the two v(CN) frequencies are separated by only 19 cm-' indicating that the stretch-stretch interaction constants are appreciably smaller for the tert-butyl isocyanide ligands than for the carbonyl ligands in accord with previous observ a t i o n ~ 'on ~ octahedral tert-butyl isocyanide metal carbonyl complexes, particularly those of the type ~ ~ C - ( M ~ , C N C ) ~ M (CY)), (M = Cr, Mo, W). In the series (Me3CNC)nW(CO)4-n(N4))I ~(n = 2-41 the )position of the de- ~ 9 v(NO) frequency ~ creases by -40 cm-' upon each successive substitution of a carbonyl group with a tert-butyl isocyanide ligand in accord with expcctations based on the weaker n-acceptor ability of tertbutyl isocyanide relative to carbon m0n0xide.l~ Colton and Commons7 have also reported some &(bident a ~ e ~ W ~ C ~ ) , ( ~ 6complexes (B)X from the ligands (C6H5I2ECH&(C6H,)2 (E = P or As) with methano bridges. We have prepared the closely related complex cis-[(CsH,),PCH2eH,p(g_:,gq,),lW(CO),(NO)I; the analogous chloride was previously prepred by a different method.15 tions of tmns-W(CO),(NO)I with appropriate reactive metallics could provide a route to novel organotungsten carbonyl nitrosyls. Thus t r a n s - ~ ( C O ) , ( N ~reacts ) ~ with thallium cyclopentadienide or, less efficiently, sodium cyclopeiitad.ienida to give the known'' CSHSW(CO)2N0,but this preparation o f CSHsW(CO)2N0is less convenient than those cur;ently in use. Unfortunately, attempts to make more interesting organotungsten carbonyl nitrosyls by variations of this method as described in the Experimental Section led eo uniformly negative results. menit. We are indebted to the National Cancer Institute for partial support of this work under Grant CAI29 38 -02. Registry NQ. cis-(Me,CNC),W(CO), (NO)I, 52699-24-8; (Me,CNC),W(CO)(NO)I, 52699-25-9; tvans-(Me,CNC),W(NO)I, 52699-26-0; [(C,H,),PCII,CH,P(C,R,),] W(CO), (NO)& 5 113229-7; tvans-W(CO),(NO)I, 39899-82-6; HW,(CO),NO, 52699-27-1. ( 1 3 ) R. B. King and M. S. Saran, Inorg. Chem., 13, 9 4 (1994). (14) F. A. Cotton, Inorg. Chem., 3, 9 0 2 (1964). (15) W. R. Robinson and M. E. Swanson, J . Organometal. Chem.,
35, 315 (1972).
Contribution No. 2525 from the Department of Chemistry, Indiana University, Bloomington, Indiana 47401
oranes. XEH.' A New
n
Jerome Rathke, David C. Moody, and Riley Schaeffer"
Received July 5,1974
AIC40434H
In this paper the synthesis and properties of Bl3pIl9 and attempts to identify intermediates in its formation are reported. A bricf report of X-ray crystal structure of this compound has already appeared.' ( I ) For the preceding paper see 3. Rathke and R. Schaeffer, Inorg. @hem., 'i 3, 760 (1974). (2) J . C . Huffman, D. C. Moody, J . W. Rathke, and R. Schaeffer, J. Chem. Soc., Chem. Commun., 308 (1973).
